Tracking of Gymnast's Limb Movement Trajectory Based on MEMS Inertial Sensor.

Area of Science:

• Sports Science
• Biomechanical Engineering
• Sensor Technology

Background:

• Accurate tracking of gymnast limb movement is crucial for performance analysis and injury prevention.
• Existing methods may lack the precision or practicality for real-time, detailed motion capture.
• Micro-Electro-Mechanical Systems (MEMS) inertial sensors offer a potential solution for unobtrusive and accurate motion tracking.

Purpose of the Study:

• To develop and validate a MEMS inertial sensor-based system for tracking gymnast limb movement trajectories.
• To assess the system's capability in recognizing different gymnastics actions.
• To evaluate the impact of sensor data errors on the accuracy of movement recognition.

Main Methods:

• A sensor network comprising MEMS inertial sensors was constructed to collect acceleration and angular velocity data from 11 body positions.
• Data preprocessing involved calculating statistical features like mean, standard deviation, information entropy, and mean square error.
• A Support Vector Machine (SVM) classification model was established using these features to recognize gymnastics movements.

Main Results:

• The system successfully recognized six different types of gymnastics movements.
• Measured acceleration values during gymnastics ranged within specific limits, with static errors accounting for only 1.6%–2% of the data range.
• The low static error indicated minimal impact on feature extraction and action recognition accuracy.

Conclusions:

• MEMS inertial sensors provide an effective method for tracking the limb movement trajectory of gymnasts.
• The proposed system demonstrates high accuracy in recognizing complex gymnastics actions.
• This technology holds promise for enhancing gymnastics training, performance evaluation, and biomechanical research.